Dehydration-Melting and Fluid Recycling during...

JOURNAL OF PETROLOGY VOLUME 38 NUMBER 9 PAGES 1255–1277 1997

Dehydration-Melting and Fluid Recyclingduring Metamorphism: Rangeley Formation,New Hampshire, USA

MATTHEW J. KOHN1∗, FRANK S. SPEAR2 AND JOHN W. VALLEY1

1DEPARTMENT OF GEOLOGY AND GEOPHYSICS, UNIVERSITY OF WISCONSIN, MADISON, WI 53706, USA2DEPARTMENT OF EARTH AND ENVIRONMENTAL SCIENCES, RENSSELAER POLYTECHNIC INSTITUTE, TROY,

NY 12180, USA

RECEIVED SEPTEMBER 3, 1996 REVISED TYPESCRIPT ACCEPTED APRIL 29, 1997

reflect the dewatering of structurally lower levels after nappe em-Muscovite and biotite dehydration-melting reactions near the peak

placement. If so, then nappe emplacement occurred at T~475°C.of metamorphism played a significant role in the reaction and fluid

history of the Rangeley Formation in southwestern New Hampshire,

USA. Evidence for in situ melting includes: (1) the consistency

among theoretical phase equilibria, observed reaction textures, and

the inferred P–T conditions; (2) disseminated, centimeter-scale,KEY WORDS: dehydration-melting; oxygen isotopes; migmatite; meta-

leucocratic quartz+plagioclase+muscovite pods; (3) diffusion andmorphic fluids; garnet zoning

growth zoning of major and trace elements in garnet that are best

explained as the result of high-T muscovite and biotite breakdown;

and (4) oxygen isotope evidence that high-T back-reaction of K-

feldspar to muscovite near peak metamorphic conditions did not INTRODUCTIONinvolve an isotopically disequilibrium (externally derived) fluid.

High-grade metapelites in New Hampshire commonlyIsotopically equilibrated fluids were apparently stored in melt pockets

exhibit retrograde muscovite that formed at the expenseand then reused as the melts crystallized, thereby driving retrogression.

of K-feldspar and sillimanite (e.g. Chamberlain & Lyons,Prograde muscovite dehydration-melting reactions further imply

1983; Thompson, 1985; Spear et al., 1990a). InasmuchP[4 kbar at TΖ~650°C, so that loading occurred before the as the prograde reaction responsible for K-feldspar inpeak of metamorphism at T~725°C. Oxygen isotope compositions these rocks has been inferred to be muscovite dehydrationof retrograde garnet that grew during cooling between T~650°C (e.g. Thompson & Norton, 1968; Chamberlain & Lyons,and T~550°C are consistent with closed-system models, indicating 1983; Thompson, 1985; Spear et al., 1990a), the back-that previous back-reaction of K-feldspar to muscovite did not disturb reaction of K-feldspar to form muscovite has commonlythe isotope compositions of the rocks. Late-stage growth of additional been proposed to result from infiltration of hydrous fluidsretrograde garnet, staurolite, and chlorite at T~475°C requires derived from dewatering of structurally lower nappesinfiltration of externally derived H2O, but this retrograde infiltration (e.g. Chamberlain & Lyons, 1983; Spear et al., 1990a;did not affect garnet and staurolite isotope compositions, as expected Spear, 1992). In southwestern New Hampshire, suchfor differing rates of infiltration-driven hydration vs isotope alteration. back-reaction is extensive in the Silurian Rangeley For-Late-stage infiltration continued after garnet and staurolite growth mation of the Fall Mountain nappe. Previous stra-ceased, as evidenced by systematic differences in isotope trends near tigraphic, petrologic, and structural studies of the nappethe base of the nappe for minerals with fast oxygen isotope diffusion (e.g. Kruger, 1946; Thompson et al., 1968; Thompsonrates (quartz, muscovite, and biotite) vs minerals with slow diffusion & Rosenfeld, 1979; Allen, 1984; Chamberlain, 1985,

1986; Spear et al., 1990a, 1995; Spear, 1992) allowed usrates (garnet, staurolite, and sillimanite). This infiltration may

∗Corresponding author. Present address: Lawrence Livermore NationalLaboratory, Earth Science, MS L-202, P.O. Box 808, Livermore, CA94551, USA. Telephone: (510)-423-8281. Fax: (510)-422-1002. e-mail:[email protected] Oxford University Press 1997

JOURNAL OF PETROLOGY VOLUME 38 NUMBER 9 SEPTEMBER 1997

to investigate in detail the nature of fluid–rock interaction Bethlehem Gneiss generally crosscuts the regional stra-tigraphy, locally, as at Bellows Falls, its contact with theduring the waning stages of metamorphism, using oxygen

isotope analysis of mineral separates and of in- metasedimentary rocks above and below is planar andbroadly conformable. The Bethlehem Gneiss also con-tracrystalline isotope zonation. As discussed below, the

new oxygen isotope data indicate that the K-feldspar– tains a well-developed foliation that is parallel to itscontacts and to the main foliation within the meta-muscovite back-reaction did not involve pervasive in-

filtration of fluids that were in equilibrium with the rocks sediments, and that is associated with the earliest phasesof Acadian deformation. Schists close to the plutonimmediately beneath the Fall Mountain nappe (the pre-

to syn-tectonic Bellows Falls pluton). We believe this contain pseudomorphs after andalusite, suggesting earlycontact metamorphism, and in some rocks the pseudo-observation eliminates two likely fluid sources: fluids

derived from the pluton itself and dewatering of struc- morphs are randomly oriented within the main foliation.These observations are most consistent with early in-turally lower nappes. Consequently, we sought an al-

ternative fluid source. trusion of the Bethlehem Gneiss (and Bellows Falls pluton)nearly parallel to the stratigraphy, with simultaneous orA reevaluation of the reaction history of the nappe rocks

revealed the previously underappreciated significance of subsequent deformation to produce the main foliation.Most importantly, peak metamorphism within the over-dehydration-melting reactions during prograde meta-

morphism. Moreover, we realized that melt segregations lying Rangeley apparently post-dated intrusion. Thus,we conclude from the regional and local geology thatprovide a local sink for prograde volatiles, which are

released upon crystallization, and are therefore available the flat contact between the Bellows Falls pluton and theRangeley Formation was present during most of theto drive retrograde reactions. Partial melting at Fall

Mountain provides the simplest explanation of many prograde and all of the retrograde metamorphism. Thegeologic relationships additionally imply that retrogradeisotope, chemical, and textural data, in that isotopically

equilibrated fluids could be stored in disseminated melt fluids could not have been derived from crystallizationof the pluton, but this possibility is further addressed bypockets during prograde muscovite breakdown, and then

recycled during cooling and crystallization to form the our stable isotope data below.retrograde muscovite. Although some retrograde chlor-ite-, garnet-, and staurolite-producing reactions did occurat T~475°C and do require infiltration of late-stage

REACTION HISTORYhydrous fluids, the proposed dehydration-melting andfluid recycling mechanism resolves long-standing issues Detailed petrologic evaluation of cation zoning and re-regarding muscovite formation and fluid budgets at high action histories is required for interpreting the oxygenT. This paper presents a revised reaction history and isotope data presented in this study. We investigated thenew oxygen isotope results, and explores the significance oxygen isotope systematics of garnet, biotite, sillimanite,of prograde melting reactions vs infiltration of externally quartz, and muscovite in different rocks, focusing onderived hydrous fluids in generating peak metamorphic different generations of slow-diffusing (‘refractory’) min-and retrograde mineralogies. erals (garnet, sillimanite, and staurolite), and isotope

zoning in refractory porphyroblastic minerals (garnet andsillimanite). We assume that refractory minerals retainthe isotope compositions of their formation, which is

REGIONAL GEOLOGY supported by the extremely slow oxygen isotope diffus-ivities indicated for garnet (Coghlan, 1990; Burton et al.,The Fall Mountain nappe is well exposed on the west

side of Fall Mountain in southwestern New Hampshire 1995; Brenan et al., 1996) and theoretical estimates forgarnet, sillimanite, and staurolite (Fortier & Giletti, 1989).(Fig. 1). The metapelite samples we analyzed are assigned

to the Silurian Rangeley Formation, and are immediately We further assume that, as they form, garnet, sillimanite,and staurolite maintain isotopic equilibrium with theunderlain by the Bellows Falls pluton. In the Fall Moun-

tain nappe, metamorphic grade reached the sil- matrix of the rock. Consequently, measuring com-positions of different generations of refractory mineralslimanite–K-feldspar zone (peak T ~725°C), and the

general P–T path is counterclockwise (early heating, and isotope zoning allows us to reconstruct the evolvingisotope composition of the rock at different times duringloading, and cooling).

The syn- and post-intrusion relationship between the metamorphism. For example, expected differences be-tween garnet cores that grew during heating and garnetBellows Falls pluton and the Rangeley Formation is

important for interpreting the stable isotope data. The rims that grew during retrogression should principallyreflect two factors: the difference in growth temperaturepluton is an outlier of the Bethlehem Gneiss, which is a

Devonian, sheet-like, felsic intrusion that is a member of garnet cores vs rims, and any open-system syn-meta-morphic changes to the isotope composition of the rock.of the New Hampshire magma series. Although the

1256

KOHN et al. DEHYDRATION-MELTING AND FLUID RECYCLING

Fig. 1. (a) Location and bedrock geology of the Fall Mountain region of southwestern New Hampshire. Axial traces of the Bronson Hillanticlinorium and Merrimack synclinorium are shown on inset map. Tick-marks show line of cross-section. (b) Cross-section of area (no verticalexaggeration). Small box shows location of detailed sketch on right of traverse from the Bellows Falls pluton (Bethlehem Gneiss) to the RangeleyFormation. Geology after Kruger (1946), Thompson et al. (1968), Thompson & Rosenfeld (1979), Allen (1984), Chamberlain (1985), Spear(1992), and Spear et al. (1990a, 1995).

1257


Fig. 2. Pressure–temperature diagram showing important reactions and P–T path. Letters correspond to specific reactions evidenced by texturesor chemical trends (see text for details); Grt1–Grt5 refer to different generations of garnet. Continuous lines are quasi-univariant reactions; dottedline shows extension of the muscovite-dehydration reaction. Shallowly inclined dashed lines are garnet molar isopleths in the anhydrousassemblage Grt+Ms+Bt+Sil+Qtz+Pl; ‘+Grt’ indicates the side on which garnet is produced. Steeply inclined dash–dot line labeled ‘+St’shows the orientation of the staurolite molar isopleth in the assemblage St+Ms+Bt+Sil+Qtz+Pl+H2O±Grt. Steeply inclined dash–dot linelabeled ‘+Grt’ shows garnet molar isopleth in the hydrous assemblage Grt+Ms+Bt+Sil+Qtz+Pl+Chl+H2O.Χ, invariant points. Locationsof the I1 and I2 invariant points and the melting reactions from Le Breton & Thompson (1988) and Spear et al. (1995) [see also Kerrick (1972),Thompson & Algor (1977), and Thompson & Tracy (1979)]. Aluminosilicate phase diagram after Holdaway (1971).

Thus, the temperature at which the refractory minerals Prograde reactionsformed and the reaction history of the rock are of (a) Chl+Qtz=Grt1+H2O. Studies of lower-grade rocksparamount importance for interpreting our oxygen iso- in the area (Spear et al., 1995) suggest an early assemblagetope data. Specifically, closed-system models that predict of Chl+Bt+Ms+Pl+Qtz±Grt1, with Grt1 having beenhow the composition of a mineral such as garnet changes produced by chlorite breakdown. In sample K92-12B,during metamorphism (Kohn, 1993; Young, 1993) can low-amplitude, patchy zoning in Ca, Y, and Sc in onebe compared with the observed variations to evaluate garnet core (Figs 3a and 4) suggests that an earlier-formedwhether open-system isotope effects are indicated (e.g. garnet (Grt1) was resorbed and/or fractured before Grt2Kohn & Valley, 1994). growth. Grt1 may not have been present in all rocks.

By reinvestigating mineral compositions and textures (b) Chl+Ms+Qtz±Grt1=Bt+And+H2O (Spear etto interpret better the stable isotope data, we have also al., 1995). This reaction formed abundant porphyroblasticrevised the reaction sequence and P–T history presented andalusite, as indicated by sillimanite pseudomorphs afterin earlier papers (Spear et al., 1990a; Spear, 1992). andalusite in the lower part of the nappe. For typicalOur discussion of the revised reaction history (Fig. 2) pelitic bulk compositions, this reaction consumes nearlyemphasizes textural and compositional features that al- all early-formed garnet (Grt1).lowed us to deduce the new reactions (Figs 3, 4, 5, and (c) And=Sil1. Abundant coarse sillimanite pseudo-6). For clarity, we denote reactions by letters and different morphs after the andalusite formed by reaction (b) in-mineral generations by numbers because textural or dicate this polymorphic transition (Rosenfeld, 1969; fig.petrologic evidence supports evidence for 12 reactions, 2A of Spear et al., 1990a; Fig. 5a).five generations of garnet, and four generations of sil- (d) Bt+Sil1+Qtz=Grt2+Ms. Grt2 occurs as cores of

larger crystals (Fig. 5b and c). It contains inclusions oflimanite. Mineral abbreviations are after Kretz (1983).

1258


Fig. 3. (a and b) Ca and Mn X-ray maps of garnet from sample K92-12B (see Fig. 5b). Patchy zoning in Ca in the garnet core may reflectovergrowth of Grt1 by Grt2. White arrows indicate inner boundaries of high-Ca Grt5 overgrowths, which preferentially grew parallel to the mica-rich foliation, and are less well developed or absent adjacent to mats of fibrolitic sillimanite (Sil4). Black arrows point towards a high-Mn ‘hump’within a few hundred lm of the garnet edge. This hump reflects high-T resorption followed by new growth of Grt4 during cooling. Such Mnhumps are ordinarily concentric to the edge of the garnet, and crosscut older zoning in Ca and trace elements. The continuous line in (b) showsthe location of the electron microprobe traverse in Fig. 4. This garnet does not contain direct chemical evidence for Grt3. Field of view is 1·5mm×1·5 mm. (c) Cr X-ray map for garnets from sample K92-12D. A well-developed high-Cr Grt3 overgrowth is indicated by white arrowsand was probably produced during dehydration-melting of biotite (Spear & Kohn, 1996). The scalloped margin of the Grt2±Grt1 core on whichthe high-Cr Grt3 grew suggests a resorption reaction immediately before melting. Grt4 and Grt5 rims are also present, but are difficult to seeusing Cr X-rays at this magnification. Scale bar at lower right corner represents ~1 mm.

1259


Fig. 4. Electron microprobe traverse across garnet from sample K92-12B, showing Mn ‘humps’ and well-developed high-Ca rims. Location ofline traverse is shown in Fig. 3b. Vertical lines separate areas corresponding to Grt1+Grt2, Grt4, and Grt5, and were drawn based on the peakin Mn (Grt1+Grt2 vs Grt4), and the increase in Ca where Mn flattens out near the rim (Grt4 vs Grt5). Compositional variations in fourth-generation garnet are only consistent with growth in the fluid-absent assemblage Grt+Bt+Ms+Pl+Qtz+Sil.

Sil1, has minor compositional zoning (Figs 3 and 4), and strong preference for K, Na and H2O cause continuouslow concentrations of Cr, Sc, and Y (Fig. 3c). These biotite (+Sil +Pl) dehydration-melting with increasingcompositions and textures are most consistent with load- T. In the Fall Mountain rocks, the increase in T to peaking after reaction (c). conditions produced garnet, as evidenced by rare, third-

(e) Grt2+Ms=Bt+Sil2+Qtz. Scalloped margins on generation garnet compositions that are preserved asGrt2 suggest partial resorption before later garnet pro- rims on older garnet cores (Spear & Kohn, 1996; Fig.duction (Fig. 3c). Nearly isobaric heating consumes a 3c). Mass balancing of minor and trace elements showssmall amount of Grt2, and causes the P–T path to pass that the garnet compositional trends are consistent withabove invariant point I1 (~4 kbar). This allows later muscovite and biotite breakdown, coupled with growthmuscovite breakdown to produce melt rather than vapor. of Grt3 during heating (Spear & Kohn, 1996). For ex-

(f ) Ms+Pl+Qtz=Sil3+melt±Kfs. This reaction ample, the micas strongly partition Cr relative to theeliminates muscovite, and produces (hydrous) melt, new other minerals. Although elimination of muscovite viasillimanite, and possibly K-feldspar. Centimeter-scale reaction (f ) does not produce garnet, the Cr con-leucocratic segregations of quartz, plagioclase, muscovite, centrations will be increased in all remaining phases.and myrmekitic quartz+plagioclase intergrowths are ubi- Subsequent breakdown of biotite via reaction (g) not onlyquitous, and commonly constitute 5–10% of the Fall causes high-Cr Grt3 to grow and to preserve a Cr-stepMountain rocks (Fig. 6a–c). Rare potassium feldspar between Grt2 and Grt3, but also continues to partitionoccurs as small inclusions or as Ζ200 lm diameter more Cr into the garnet, because high-Cr biotite is beingdomains in large plagioclase grains. Clots of fibrolitic consumed. Thus, reactions (f ) and (g) should producesillimanite surround leucocratic segregation (Fig. 6d) and garnet that has a much higher Cr content than previouslymay be prograde reaction products. We interpret the formed garnet, and whose Cr content increases as itleucosomes as former pockets of melt produced during grows (i.e. as observed in Fig. 3c).reaction (f ) that subsequently cooled to produce newquartz, plagioclase, and muscovite via reactions (h) and(i).

Retrograde reactions(g) Bt+Sil1–3+Pl+Qtz=Grt3+Kfs+melt. After mus-(h) Kfs+Grt2–3+(hydrous) melt=Bt+Sil4+Pl+Qtz.covite dehydration-melting, reaction progress is stronglyThis is the reverse of reaction (g). Second- and third-controlled by the thermodynamic and chemical prop-

erties of the melt. The large entropy of the melt and its generation garnet are replaced by coarse biotite grains

1260


Fig. 5. Retrograde mineral textures. (a) Sample K92-12G. Box shows area imaged in (d). A large sillimanite pseudomorph after andalusite (Sil1) hasbeen altered to fine-grained muscovite (Ms) along its margins. Small idioblastic crystals of staurolite (St) are common in these muscovite rinds,suggesting the reaction Sil+Bt+H2O±Grt=St+Ms. (b) Sample K92-12B. Fibrolitic sillimanite (Sil4) abuts and replaces garnet. Second-generationgarnet core (Grt2) has fifth-generation garnet overgrowths (Grt5), which preferentially grew outward within the foliation plane and contain inclusionsof Sil4. X-ray maps and a compositional traverse for this garnet are also shown in Figs 3 and 4. (c) Sample K92-12B. A large Grt2 core is overgrownon one end by a thick rim of Grt5. (Note abundance of ilmenite inclusions in Grt5 and elongation of Grt5 within the micaceous foliation.) (d) Close-up of area from sample K92-12G, showing staurolite crystals and fifth-generation garnet rims. Grt5 rims are well developed on earlier cores (Grt2),and have a characteristic increase in ilmenite inclusion density. Scale bars in lower left corners all represent 1 mm.

and are abutted, surrounded, and replaced by mats of leucosomes and the matrix as originating from this re-action, although the matrix muscovite may well havefibrolitic sillimanite [Fig. 5b; see also fig. 2B of Spear et

al. (1990a)]. Resorption almost completely removed Grt3 changed chemical composition because of later reactions.X-ray maps of leucosomatic muscovite grains (M. J.and caused an increase in Mn towards the rim of the

garnet (e.g. Spear & Florence, 1992). The fibrolitic sil- Kohn, unpublished data, 1997) show higher Ti in corescompared both with rims and with matrix muscovitelimanite clots observed on leucosome margins may have

formed by this reaction, rather than by reaction (f ). grains, consistent with initial nucleation at high T. Vari-ations in the Ti content and Fe/(Fe+Mg) of biotite were(i) Sil1–4+(hydrous) melt±Kfs=Ms+Pl+Qtz. This

reaction is the reverse of reaction (f ), and allowed for- also described by Spear et al. (1990a), and ascribed toretrograde reaction and diffusional exchange.mation of high-T retrograde muscovite and the near-

elimination of any K-feldspar. Micas in the leucosomes ( j) Bt+Sil1–4+Qtz+Pl=Grt4+Ms. Fourth-gen-eration garnet compositions are manifested by a rimwardare oriented randomly relative to the foliation (Fig. 6c),

and fibrolitic sillimanite in the matrix and on the margins decrease in Mn from a ‘hump’ near the outer margin oflarge garnets (Figs 3b and 4). The Mn hump crosscutsof leucosomes is typically surrounded by coarse-grained,

cross-cutting muscovite (e.g. Spear et al., 1990a, fig. 2D; Ca and trace element zoning in earlier-generation garnetsand is concentric about scalloped margins (Fig. 3b).Fig. 6d). We view the coarse muscovite in both the

1261


Fig. 6. Melting textures. (a) Outcrop photograph showing distribution of cm-scale leucocratic segregations. (b) Sample K95-18D. Photomicrographin plane-polarized light of two leucosomes, a sigmoidal one above and a larger composite one below. ‘Leuc’ indicates different leucocratic bodies.Boxes show areas of (c) and (d). (c) Inverse silicon X-ray map of the upper leucosome in Fig. 3b. Quartz is dark, plagioclase is intermediategray, and muscovite, sillimanite, and biotite are light gray. Coarse muscovite in the center of the leucosome is randomly oriented and intergrownwith quartz. Feldspar and quartz along the leucosome margins show myrmekitic intergrowths. (d) Aluminum X-ray map showing fibroliticsillimanite (small white specks) surrounded by muscovite at the boundary between two leucosomes. Sillimanite in the matrix is surrounded bycoarse muscovite in a similar texture. The fibrolitic sillimanite and the leucosomes are interpreted as the products of muscovite dehydration-melting, whereas the muscovite is believed to have overgrown the sillimanite during later melt crystallization upon cooling. Scale bars in lowerleft corners all represent 1 mm.

These humps are probably the result of early resorption 2A and D of Spear et al. (1990a)]. This reaction requiresa small amount of externally derived fluid.[reaction (h)] followed by regrowth (Spear et al., 1990a,

1995; Spear & Florence, 1992). Grt4 also contains in- (l) Sil1–4+Bt+H2O=Grt5+Chl+Ms+Qtz. In manysamples retrograde chlorite that contains fine-grainedclusions of sillimanite, and shows abrupt increases in Sc

and Y, and decreases in Ca and Cr compared with Grt3 inclusions of ilmenite makes up ~1% to ~30% of themode; it has replaced biotite. Grt5 rims are identifiable(Spear & Kohn, 1996). The trace and major element

composition trends and the fact that garnet is growing by an increase in the number of fine-grained ilmeniteinclusions (Spear et al., 1990a, fig. 2B and D; Fig. 5b–d),at all are only consistent with production of Grt4 in the

muscovite stability field (Spear et al., 1990a). a pseudomorphic texture after coarse-grained matrixbiotite, and an abrupt increase in Ca followed by a(k) Sil1–4+Bt+H2O±Grt2–4=St+Ms+Qtz. In a few

samples, staurolite occurs in muscovite+staurolite decrease to the garnet edge (Figs 3 and 4). Mn remainsroughly constant, Fe/(Fe+Mg) and Y increase, and Cr,pseudomorphs after sillimanite [Fig. 6a and d; also fig.

1262


Sc, and P decrease. The high-Ca rims grew preferentially are expected to produce K-feldspar (e.g. Thompson &Algor, 1977). In either case, the amount of K-feldsparin the plane of the main foliation. In one sample, elongateproduced is far less than if the muscovite dehydrationGrt5 rims contain small inclusions of staurolite, althoughreaction is crossed, and thus melting probably betterGrt5 also occurs in rocks that lack staurolite. A moreexplains the limited occurrence of K-feldspar.detailed justification of the reaction responsible for form-

Spear et al. (1990a) and Spear & Florence (1992)ing chlorite and Grt5 is presented in Appendix A.assumed the high-T production of garnet and K-feldsparin the Fall Mountain nappe resulted from dehydrationreactions, rather than melting. Both interpretations ex-

Further considerations of melting plain the mineral assemblages and garnet zoning, becauseSeveral criteria have been proposed to evaluate whether both rely on muscovite and biotite breakdown to produceleucocratic segregations are indicative of partial melting K-feldspar and garnet. However, melting better explains(e.g. McLellan, 1983, 1989; Ashworth & McLellan, 1985), the cm-scale leucocratic segregations, the observed re-including appropriate peak P–T conditions, large grain hydration without concomitant isotope effects (as dis-size, random mineral orientation and distribution, and cussed below), and the paucity of K-feldspar. Thesesimilarity of bulk composition to expected melt com- revisions affect the prograde P–T path: the ‘melting’positions. As demonstrated by Spear et al. (1990a), the model requires that the prograde path pass above the I1peak P–T conditions are consistent with the production invariant point (~650°C and 4 kbar; Fig. 2).of partial melts via the muscovite dehydration-melting Because melts produced by dehydration-melting re-reaction, and Fig. 6 shows that the grains within the actions such as (g) are strongly undersaturated withsegregations are randomly oriented and 5–10 times larger respect to H2O (e.g. Burnham, 1979), the water producedthan matrix minerals. Modal analysis using X-ray maps from mica breakdown simply dissolves into the melt.indicates that the bulk compositions of the leucosomes During subsequent cooling in a closed system, the sameare encompassed by the large range of melts produced reactions are crossed in the reverse direction, andexperimentally by mica dehydration-melting (e.g. Le muscovite±biotite±sillimanite reappear and grow untilBreton & Thompson, 1988; Vielzeuf & Hollaway, 1988; the hydrous melt is used up. At that point, a rock wouldPatino-Douce & Johnston, 1991; Gardien et al., 1995), contain the ‘rehydrated’ assemblage Grt+Bt+with the exception that the water content of the leuco- Ms+Sil+Qtz+Pl, but the back-reaction to producesomes (<2 wt %) is substantially lower than found in muscovite did not involve infiltration. Partial melting canexperiments (4–10 wt %). This is consistent with our thus allow fluids to be stored in melts at high temperaturehypothesis that water from the melt also reacted with and recycled during cooling to drive retrograde reactions.

Further cooling then allows the water-conserving reactionmatrix minerals such as K-feldspar and sillimanite toBt+Sil+Qtz+Pl=Grt+Ms to proceed.produce coarse matrix muscovite. The distribution of

minerals within the leucosomes is not random (e.g. mus-covite is concentrated in the centers, and myrmekiticquartz and plagioclase near the margins), but if meltscrystallize inhomogeneously during cooling, then the

OXYGEN ISOTOPE DATAdistribution of minerals within the leucosomes should bed18O of bulk mineral separates andinhomogeneous. For example, if initial crystallization ofdifferent mineral generationsquartz and plagioclase occurred on leucosome margins,

then the mineral that nucleated last (muscovite) would Garnet, staurolite, biotite, sillimanite, muscovite andform in the centers of the leucosomes, as observed (Fig. quartz from finely ground samples of the Rangeley For-6). Thus, many lines of evidence are consistent with a mation, and quartz, feldspar, muscovite, biotite, andpartial melt origin for the leucosomes. garnet from samples of the Bethlehem Gneiss were sep-

The modal amount of melt is limited by the modal arated and analyzed for their oxygen isotope compositionsabundance of coarse muscovite in the leucosomes and (see Table 2, in Appendix B). In several Rangeley For-matrix. Based on the present mode of coarse muscovite, mation samples, two different populations of garnet werewe predict the proportion of melt produced to be ~5– readily distinguished. The first is nearly inclusion free,15%, similar to the observed abundance of leucosomes. ‘bubble-gum pink’ in color, and has no crystal faces,The amount of K-feldspar produced by melting reactions whereas the second contains a small amount of fine-is less clear. For example, Patino-Douce & Johnston grained ilmenite inclusions, is orange, and ordinarily has(1991) found no evidence for K-feldspar in their melting well-developed crystal faces. A few pink garnet fragmentsexperiments on a natural muscovite-bearing metapelite have orange rims. For these Rangeley samples, thin-at their lowest run conditions (825°C, 10 kbar), whereas section observations allow different garnet populations

(pink vs orange) to be correlated with petrologicallymost simplified muscovite dehydration-melting reactions

1263


distinct garnet generations (Grt1–5). The pink population d18O zoningwith few inclusions and no crystal faces mainly represents Only garnet and sillimanite are sufficiently coarse grainedGrt2. Although it could also contain Grt1 and Grt3, our to permit the measurement of isotope zoning profiles.X-ray maps show these generations are volumetrically We found no correlation between isotope compositioninsignificant. The orange population with crystal faces and analysis location within coarse sillimanite in anyand fine-grained ilmenite inclusions is readily correlated sample, nor from garnet from samples K92-12C and

K92-12A (Fig. 8a). In contrast, analyses of two garnetswith Grt4 and Grt5. In Bethlehem Gneiss samples, garnetfrom sample K92-12D, collected within 50 cm of thegrains are ubiquitously rounded, suggesting that they areRangeley–Bethlehem Gneiss contact, allow us to derivexenocrysts from assimilated schist. If they are xenocrysts,a composite profile that exhibits a dramatic decrease inthen given the extremely slow diffusivity of oxygen ind18O from ~12·3‰ to ~11·5‰ within ~400 lm of thegarnet (Coghlan, 1990; Burton et al., 1995; Brenan et al.,garnet edge (Fig. 8b). This outer region corresponds to1996), they may preserve an earlier d18O signature, outthe Mn ‘hump’ (Grt4) and high-Ca overgrowths (Grt5)of equilibrium with the rest of the gneiss.observed in the cation zoning. Although the Grt5 rimsMineral separate data are plotted in Fig. 7 vs distanceon the coarse K92-12D garnet porphyroblasts are toofrom the Bethlehem–Rangeley contact. Garnet and sil-small to sample directly, small Grt5 grains were separatedlimanite within the Rangeley >1 m from the contactfrom different layers within 5 cm of the large garnets,show small variations in d18O that are probably pre-and yield d18O values of ~11·2‰. Thus, there is ametamorphic (sedimentary or diagenetic), as well as muchsystematic decrease in d18O from second- to fifth-gen-

greater variation between 25 and 50 cm of the contact. eration garnets in the sample, similar to the second- vsThere is additionally an abrupt isotope shift from the fifth-generation garnet d18O decreases observed in otherRangeley into the Bethlehem. This shift is not the result rocks (Fig. 7a).of differences in mineralogy or mineral fractionations.Although aqueous fluids are metastable at T[~650°Crelative to granitic melts, the d18O of a hypothetical fluid

INTERPRETATIONS(or any other phase) at peak conditions (725°C) can beTiming of formation of Qtz+Ms+Btcalculated from measured compositions. For this cal-isotope profilesculation, we used compositions of prograde garnets from

the Rangeley because of the possibility that retrograde The composition vs distance trends measured for fast-hydration affected compositions of other minerals, and diffusing minerals (quartz, muscovite, and biotite) are

clearly distinct from those of the slow-diffusing mineralswe used the compositions and modes of the matrix(Grt2 and Sil; Fig. 7). For example, the expected com-minerals of the Bethlehem Gneiss (see Table 2, in Ap-position trend for quartz based on measured Grt2 andpendix B) because we were unsure whether the garnetssillimanite compositions (dotted line, Fig. 7b) differs sig-were isotopically disequilibrium xenocrysts. The hy-nificantly from the measured trends exhibited by quartz,pothetical fluid compositions are shown by filled squaresmuscovite, and biotite. This implies that the decrease inin Fig. 7a, and show a similar compositional shift. In-quartz, muscovite, and biotite d18O in Rangeley samplesterestingly, the isotope compositions of Bethlehem Gneisswithin 15 m of the Bethlehem Gneiss must have occurredgarnets are within uncertainty in equilibrium with co-after the formation of at least Grt2 and coarse sillimanite.existing quartz, feldspar, and mica, as recalculated forAlthough less obvious, the measured quartz, muscovite,the peak of metamorphism, and yet very different fromand biotite profiles also cannot be reconciled with theearly-formed garnets in the Rangeley Formation. Thiscompositions of retrograde Grt4+Grt5 and staurolite. Ifsuggests that either a schist other than the Rangeleyretrograde garnet formed after isotope alteration of

supplied these garnets as xenocrysts, or that the garnets quartz, muscovite, and biotite, then the expectedin the Bethlehem Gneiss recrystallized. All but one sample Grt4+Grt5 composition (dotted line, Fig. 7a) would beof Grt4+Grt5 show a decrease of 0·1–0·6‰ relative to 1–2‰ lower than observed compositions in nearly allGrt2 (Fig. 7a). Quartz, muscovite, and biotite show much rocks within 15 m of the contact. Furthermore, as dis-smoother compositional trends, significantly different cussed below, most of the measured staurolite com-from the small-scale variations and steep isotope gradients positions indicate crystallization before any whole-rockobserved for garnet and sillimanite (Fig. 7b). The dotted isotope alteration. These observations imply that thelines in Fig. 7 allow comparison of the observed isotope decrease in d18O for fast-diffusing minerals occurred aftervariations for quartz vs garnet. The disparities between formation of Grt4+Grt5 and staurolite (i.e.these trends suggest that isotope exchange or transport Ζ450–500°C), and so the isotope alteration reflects aoccurred at the base of the Fall Mountain nappe after late-stage process, unrelated to the formation of high-T

muscovite at T[650°C.crystallization of the refractory minerals.

1264


Fig. 7. Oxygen isotope compositions of mineral separates vs distance from the contact between the Bethlehem Gneiss and the RangeleyFormation. Continuous lines were drawn by eye to roughly fit the observed data trends for early garnet and for quartz. (Note vertical scalechange between diagrams.) (a) Measured compositions of slow-diffusing Grt, St, and Sil, and calculated composition of a hydrous fluid (had onebeen present) at peak metamorphic conditions. Open circles with diagonal rule show Grt2±Grt1±Grt3 compositions in rocks for which staurolitecompositions were also measured. Symbols for St were offset slightly to facilitate comparison. The dashed line shows the predicted compositionof Grt4+Grt5 rims based on the measured compositions of Grt2±Grt1±Grt3 and a closed-system model; the dotted line shows predictedcomposition of Grt4+Grt5 based on the measured Qtz profile and fractionations observed in rocks far from the contact. The compositional shiftat ~50 cm from the contact and small-scale variations in isotope composition at 5 and 17 m suggest little net transport of oxygen at a meter-scale before or during Grt2±Grt1±Grt3 and Sil formation. The similarity between the compositions of retrograde St and Grt2±Grt1±Grt3 forthree of four samples, and the closer correspondence of measured Grt4+Grt5 rims with the dashed line rather than the dotted line are bothconsistent with closed-system models. This correspondence implies that alteration of Qtz, Ms, and Bt compositions post-dated growth of Grt4,Grt5, and St. (b) Measured compositions of fast-diffusing Qtz, muscovite, and biotite. The dotted line shows expected Qtz values based oncompositions of Grt2 and Sil, and observed fractionations in rocks far from the contact. The smoother trend and 18O depletion in the RangeleyFormation 0–10 m above the contact suggest that net transport and/or exchange of oxygen occurred before the closure temperatures of theseminerals (T[275–450°C). Measurement uncertainties are equal to or smaller than the symbol sizes.

1265


Fig. 8.

As described by Kohn (1993) and Young (1993), com-Isotope modeling of closed-systembining isotope partitioning and mass balance equationsvariationsallows prediction of isotope changes and mineral isotopeOne important question is whether changes in whole-zoning that accompany metamorphic reactions androck isotope compositions are required by the measuredchanges of P and T (i.e. within a closed system). Bycompositions of retrograde garnet and staurolite. Insofarspecifying ‘unaltered’ initial compositions (based on meas-as garnet and staurolite form in equilibrium with theured Grt2) and the reaction and P–T sequence, therock and are immune to diffusional resetting, their com-compositions of all minerals throughout the metamorphicpositions will reflect whole-rock compositions at the timeevolution can be predicted. These predicted isotope com-they formed. However, even in a non-infiltrated rockpositions depend on mineral fractionations, modal abund-(i.e. closed-system), different generations of a mineralances, changes of mineral assemblage, and fractionalsuch as garnet might not have identical compositions.crystallization and fluid distillation processes (Kohn,Instead, the composition of a specific garnet generation 1993; Young, 1993). However, for most rocks, mineral

will depend on the prior reaction history, bulk com- isotope compositions depend essentially only on T. Forposition, and the temperature at which that generation example, a T increase of 10°C increases garnet d18O bygrew (Kohn, 1993; Young, 1993). Therefore, detailed 0·04–0·1‰ in typical metapelites (Kohn, 1993). No singlemodeling of closed-system isotope variations is required garnet in the Rangeley Formation preserves the entireto determine whether the lower d18O measured for reaction history, but a P–T diagram contoured for garnetGrt4+Grt5 compared with Grt2 reflects formation at d18O taking into account the complex reaction historydifferent temperatures, or instead reflects infiltration by and an idealized profile for a hypothetical garnet thatlow d18O fluids. Retrograde garnet+staurolite and pro- preserves Grt1–5 (Fig. 9) allow the following predictionsgrade Grt2 are most important because their growth to be made:brackets formation of high-T muscovite. If closed-system (1) Garnet cores ideally should show a rimward increasemodels accurately predict the compositional differences in d18O of ~1‰ corresponding to a DT of ~200°Cobserved among Grt2, Grt4+Grt5 and staurolite, then (Grt1 to Grt3). However, because only Grt2 is typicallyformation of high-T muscovite could not have involved preserved, isotope zoning in garnet cores will more likelymajor oxygen isotope alteration of the rocks. If instead, be flat, with possible isolated analyses of low and highmeasured Grt4+Grt5 and staurolite d18O values are lower d18O because of relict Grt1 and Grt3.than predicted by closed-system models, then the differ- (2) Grt4 should show a decrease in d18O of 0·0–0·5‰ence between predicted and measured values may allow relative to Grt2 garnet cores, as a result of the decreasethe degree of isotope alteration by externally derived in temperature (from 650 to 550°C) in the assemblage

Grt+Bt+Sil+Ms+Qtz+Pl.fluids to be characterized.

1266


Fig. 8. (a) Oxygen isotope zoning profile across a garnet from sample K92-12A, showing constant d18O. Inset shows location of the rim-to-rimtraverse. (b) Sketch and composite plot of oxygen isotope zoning data from two garnets in sample K92-12D, showing strong decrease in d18Otowards garnet rims. This decrease is consistent with late garnet growth at much lower temperature than the garnet core. Dashed lines indicatelocation of saw cuts used to dissect sample.

(3) If Grt5 formed at T~475°C, its d18O should be at the isotope data indicate that the rocks were isotopicallyleast 0·5‰ lower than Grt2 garnet cores. unaltered between growth of Grt2 and growth of Grt5

(4) Assuming retrograde staurolite formed at T~475°C, and staurolite. This implies that back-reaction of K-its d18O value should be ~0·1‰ lower than Grt2 [D(St– feldspar to muscovite did not substantially affect whole-Grt)~0·6‰; Kohn & Valley, 1997]. rock isotope compositions, and that the outcrop-scale

Our data match the closed-system predictions fairly profiles measured for quartz, muscovite, and biotite (Fig.well. The absence of oxygen isotope zoning in garnet 7b) were produced after formation of Grt5 and staurolite.cores is consistent with the preponderance of Grt2 andpoor preservation of Grt1 and Grt3. The systematic de-crease in d18O from garnet cores to Grt5 is consistentwith growth of the later generations of garnet during

Sources of high-T retrograde fluidscooling. Three of the four separates of retrograde stau-Data collected from many rocks across the Bethlehem–rolite show d18O values similar to or slightly lower than

Grt2, as expected from closed-system models. Therefore, Rangeley contact demonstrate the difficulty in reconciling

1267


Fig. 9. P–T diagram contoured for garnet d18O assuming reference conditions of 12·4‰ (V-SMOW) at 600°C and 5 kbar, with the preferredP–T path and locations of different garnet generations. Inclined dashed lines indicate locations of major assemblage changes that resulted ineither garnet growth or garnet consumption. Inset shows schematic zoning profile for a hypothetical garnet that preserves all five generations.In the profile, continuous lines represent well-preserved garnet generations (Grt2, Grt4, and Grt5), and dashed lines represent rarely preservedgenerations (Grt1 and Grt3). Staurolite would have a composition ~0·6‰ higher than garnet.

the oxygen isotope data with an origin of high-T ret- no evidence of such an infiltration front and are, in fact,consistent with closed-system equilibration. Even Grt5rograde muscovite as the product of open-system hydrous

infiltration and reaction after K-feldspar, as was inferred and staurolite, which are interpreted to have formed atT<500°C, show no isotopic shift from that predictedpetrographically in previous studies. The thickness of the

Fall Mountain klippe is at least 500 m, and the average from a closed system, and so could not have formed inan isotopically altered rock. Although quartz, muscovite,mode of coarse muscovite in these rocks is ~10%. Pro-

duction of 10% muscovite solely by infiltration-related and biotite do show isotope evidence of infiltration nearthe contact, this alteration must have occurred at low Tconversion of K-feldspar requires ~1 mole of H2O for

each 100 moles of oxygen in the rock. For a section of 500 after the formation of Grt5 and staurolite. We thereforeconclude that retrogression of K-feldspar did not involvem thickness, assuming that fluid flow was perpendicular to

the contacts, and assuming that the concentration ratio either (1) fluids derived from reactions or crystallizationin the Bethlehem Gneiss itself, or (2) fluids derived fromof oxygen between fluid and rock is 1·6 (Baumgartner &

Rumble, 1988), production of this volume of muscovite other sources (e.g. underlying nappes) that flowed throughand equilibrated with the gneiss.implies a minimum fluid flux of ~800 cm3/cm2. For a

fluid flux of 800 cm3/cm2, instantaneous equilibration In contrast to fluid infiltration, we propose anatexis andsubsequent melt crystallization as the simplest explanationbetween fluid and 80% of each rock implies an isotope-

front advection distance of ~10 m (Baumgartner & for the prograde Sil+Kfs-grade dehydration and ret-rograde rehydration of the Rangeley Formation. If mostRumble, 1988). At the temperatures at which retrograde

muscovite formed (~650°C; Spear et al., 1990a), any or all of the water associated with high-T muscovitebreakdown were stored in in situ melts, then melt crys-infiltrating fluid will equilibrate isotopically within tens

of thousands of years with the fast-diffusing minerals tallization upon cooling would simply use that water toproduce muscovite, without a significant isotope effect.(quartz, feldspar, and micas), which modally constitute

[80% of the Rangeley schists. This would then result in a rehydrated rock with noconcomitant large isotope front or evidence for open-Measured oxygen isotope compositions of the re-

fractory minerals garnet, staurolite and sillimanite show system behavior during retrograde garnet growth, in

1268


better agreement with the observed isotope variations. Our isotope data strongly support the hypothesis thatlate-stage hydration was driven by infiltration of fluidsFor a typical volumetric abundance of leucosomes inthat had first equilibrated with the Bethlehem Gneiss.outcrops (5–10%) and a typical water content of a felsicThe consistency of the isotope compositions of Grt5 andmelt produced by dehydration-melting of muscovite (~10staurolite with a closed-system model is expected fromwt % H2O; e.g. Le Breton & Thompson, 1988), there isdifferences in the propagation rates for the hydration vssufficient water stored in the melts to produce the ~10%isotope alteration fronts. The hydration of the lowest 50coarse muscovite that is observed in the leucosomes andm of the nappe to produce 2% chlorite advances anmatrix.oxygen isotope front only ~1

2 m, and chlorite, staurolite,One additional phase equilibrium consideration prefersand Grt5 at the nappe base should have formed longour ‘melting’ interpretation over an infiltrative mech-before the oxygen isotope values of their constituentanism. As described by Spear et al. (1990a), the com-whole rocks were affected. Thus, Grt5 and staurolite grewpositional changes in Grt4 are only consistent with growthas the hydration front passed, but stopped growing beforeduring cooling in the fluid-absent assemblagethe oxygen isotope front reached them, and so containGrt+Sil+Ms+Bt+Qtz+Pl, through the reactionno evidence for infiltration in their d18O values. InSil+Bt+Qtz+Pl=Grt+Ms. Specifically, for this as-contrast, minerals with fast oxygen diffusivities shouldsemblage and reaction, the Ca content of the garnethave continued to change isotope composition as in-changes very little. The presence of a hydrous fluid duringfiltration produced chlorite higher in the nappe, andretrogression substantially changes mineralogy and com-record an isotope front of several meters. This explainspositions, however, because it drives the simultaneouswhy quartz and mica are isotopically altered many metersreaction: Pl+Sil+H2O=white mica+Qtz. This re-from the contact (Fig. 7b): their closure temperaturesaction consumes the albite component of plagioclase,with respect to oxygen diffusion ([450°C for quartz andand after cooling of 50–100°C (Spear et al., 1990a) should~300°C for muscovite and biotite; Giletti & Yund, 1984;produce paragonite and extremely anorthitic plagioclaseFarver & Yund, 1991; Fortier & Giletti, 1991) are at or(~An70). To maintain partitioning with such anorthiticbelow the formation temperature of Grt5 (~475°C=plagioclase, Grt4 should contain high grossular contents.infiltration T ).The absence of both paragonite and high Ca in Grt4

implies that hydration was not an important processduring cooling and Grt4 growth (i.e. after the high-Tmuscovite was produced). This is expected if the fluids

DISCUSSIONwere derived from in situ melts, because rehydration haltsAlternative hypotheses for high-Tafter the melt has crystallized. However, if externallyretrograde muscovitederived fluids were responsible for hydration, their in-

filtration must have ceased immediately after the K- Alternative hypotheses of rehydrating the Fall Mountainfeldspar to muscovite reaction. Such fortuitousness seems nappe by exotic fluids at high T to produce coarseunlikely. retrograde muscovite must simultaneously address fluid

source and transport. After considering in detail fluidproduction by dehydration of either the same rocks at

Origin of late hydrous phases greater depth or isotopically similar schists from the nextAlthough no external fluid is required to produce the lower nappe, and fluid transport via fractures or alonghigh-T retrograde muscovite, the low-T growth of stau- foliation planes, we conclude that infiltration by hydrousrolite, chlorite (and probably Grt5) does require late fluids at high T was improbable.infiltration of hydrous fluids. As the most likely source One possible source of high-T fluids is from deeperof late-stage fluids is from structurally lower nappes, we levels of the nappe itself, either via prograde dehydrationassume that reaction and equilibration of the Rangeley reactions or retrograde crystallization of melts. Suchwith these exotic fluids would have occurred pro- fluids would have been in isotope equilibrium with thegressively: the base of the nappe was probably affected Rangeley Formation, and if capable of moving to ourfirst, and infiltration-driven mineralogical and isotopic sample location would have produced little or no isotopechanges then swept upward into the nappe. Production effect. However, it is unlikely such fluids existed at theof 2% chlorite on average in the exposed ~500 m section time the muscovite formed. At the high T reached byof the nappe requires a fluid flux of at least 440 cm3/ the nappe, a free fluid is metastable with respect to melt,cm2. If these fluids equilibrated with the Bethlehem Gneiss and so prograde reactions deeper within the nappe wouldand chromatographic theory applies (Baumgartner & have produced melt rather than a mobile hydrous fluid.Rumble, 1988), then infiltration to produce the observed Furthermore, at solidus to sub-solidus temperatures, alate-stage hydration should have advanced a low d18O hydrous fluid will react first with K-feldspar to produce

muscovite and then with plagioclase to produce sodicisotope front several meters into the nappe.

1269


mica. The nappe does not exhibit evidence either for the metamorphic peak that is observed in high-grademetapelites from New England (e.g. Chamberlain &voluminous melts or for back-reaction of plagioclase,

implying that there was insufficient hydrous fluid to both Lyons, 1983; Thompson, 1985; Spear et al., 1990a).Because many of these rocks are locally migmatitic, theoverwhelm the rehydration capacity of the source rocks

and additionally cause rehydration of rocks higher up in fluids required to produce this muscovite may have beenobtained from in situ melts rather than by infiltration.the nappe.

As proposed by Spear et al. (1990a), hydrous fluids In contrast, K-feldspar-bearing assemblages are pre-served in other terranes, and this requires loss of thederived from metapelites of the underlying, lower-T

Skitchewaug nappe (Fig. 1b) might have percolated up- fluid derived from muscovite (±biotite) breakdown. Forexample, in metapelitic gneisses of the Chesham Pondward and retrogressed the overlying Fall Mountain

nappe. One fluid-flow mechanism might be through nappe, which structurally overlies the Fall Mountainnappe (Fig. 1b), K-feldspar porphyroblasts are commonfractures, allowing fluid to traverse the Bethlehem Gneiss

and disperse itself within the Fall Mountain nappe without (e.g. Thompson, 1985; Spear, 1992), and many leuco-some-bearing samples contain no late cross-cutting mus-prior equilibration with the Bethlehem Gneiss. If the

fluid was derived from metapelites that were isotopically covite. These mineralogical differences may reflectdifferences in metamorphic pressures. The Cheshamsimilar to the Rangeley Formation, such rapid transport

might allow rehydration without producing a cor- Pond nappe was metamorphosed at lower pressure(Spear, 1992), and it is likely that its P–T path passedresponding isotope front and without significantly affect-

ing the bulk-rock isotope compositions. This possibility below the I1 invariant point (Fig. 2), causing muscoviteto dehydrate to produce K-feldspar and sillimanite beforeseems unlikely, given the high ductility of upper am-

phibolite-facies quartzo-feldspathic rocks and pelitic any melting. Loss of that water then prevented back-reaction of the K-feldspar to form muscovite duringschists, and the spatially uniform production of high-T

retrograde muscovite, especially inside leucosomes. On cooling. In other rocks, back-reaction may be limited bypooling or extraction of melt, which increases the lengththe contrary, we would expect fracture-flow to con-

centrate fluids in larger channels, rather than disperse scale over which melts and minerals would be requiredto communicate. For example, in K-feldspar-bearingthem. Alternatively, if fluids from the Skitchewaug nappe

traversed and equilibrated with the Bethlehem Gneiss, migmatitic rocks from Massachusetts (Tracy, 1978) andNew Hampshire (Thompson, 1985), melt segregationsbut then flowed along the foliation to our sampling

location, they might have already equilibrated with the are much larger than the ~1 cm leucosomes observedat Fall Mountain. Crystallization of large melt pools couldRangeley, causing no isotope effect in garnet or staurolite.

However, it is unclear why fluids would enter the Fall produce local retrograde muscovite after sillimanite, butefficient communication between the segregated meltMountain nappe only ‘upstream’ of our samples and not

at the sampling location. Furthermore, at the point of and dispersed K-feldspar might be difficult. In rocks inwhich K-feldspar is the product of dehydration-melting,infiltration such fluids should additionally have caused

formation of paragonitic micas and anorthitic feldspars, K-feldspar should be retained in those rocks not in directcontact with melt or from which the melt escaped.which have not been observed.

In summary, field, textural, and isotope data do notsupport the hypothesis that high-T retrogression of K-feldspar to muscovite was the result of infiltration of

Tectonic implicationshydrous fluids. Instead, the observations are best ex-If the Rangeley Formation underwent muscovite de-plained by anatexis and subsequent crystallization of inhydration-melting, then the rocks must have reached asitu melts.pressure of at least 4 kbar by the time they attained atemperature of ~650°C (to pass above invariant pointI1 in Fig. 2). That is, most of the loading preceded

Implications for fluid budgets the thermal maximum. This interpretation is furthersupported by the late-stage metamorphic evolution ofAn important implication of the combined petrologic and

stable isotope analysis is that retrograde fluid recycling the Fall Mountain rocks. All the zoning trends andmineral textures are consistent with a simple retrogradecan occur in rocks that have experienced dehydration-

melting. Fluids produced by mica dehydration-melting history involving substantial cooling with little or noexhumation. The only break in this history involvesmay be stored in melt pockets and become available for

back-reaction during melt crystallization on cooling (e.g. late-stage fluid infiltration (450–500°C) to produce theretrograde staurolite, chlorite and Grt5, as well as theAshworth & McLellan, 1985; Olsen, 1987). This model

explains the common occurrence of coarse, late muscovite 18O depletions in muscovite, biotite, and quartz isotopecompositions (Fig. 7b) close to the contact. We believethat was produced at the expense of sillimanite soon after

1270


this fluid was derived from dewatering of less strongly REFERENCESmetamorphosed Skitchewaug nappe (Fig. 1b), which im- Allen, T., 1984. The Fall Mountain outlier: a piece of the Fall Mountainplies that final emplacement of the Fall Mountain nappe nappe. B.A. Thesis, Harvard University, Cambridge, MA.occurred at TΖ500°C. Ashworth, J. R. & McLellan, E. L., 1985. Textures. In: Ashworth. J.

R. (ed.) Migmatites. New York: Chapman and Hall, pp. 180–203.Baumgartner, L. P. & Rumble, D. R., III, 1988. Transport of stable

isotopes: I: Development of a kinetic continuum theory for stableCONCLUSIONS isotope transport. Contributions to Mineralogy and Petrology 98, 417–430.

Bowman, J. R., Willett, S. D. & Cook, S. J., 1994. Oxygen isotopeDifferent petrologic techniques can clearly facilitate quan-transport and exchange during fluid flow: one-dimensional modelstitative interpretation of tectonism and metamorphism.and applications. American Journal of Science 294, 1–55.Our initial isotope data for garnet, which suggested that

Brenan, J. M., McKeegan, K. D. & Ryerson, F. J., 1996. Oxygenlittle fluid advection had occurred across the Bethlehemdiffusion in garnet at high pressure: experimental technique and

Gneiss–Rangeley contact, prompted us to investigate initial results. EOS Transactions of the American Geophysical Union 77,anatexis as a source of high-temperature retrograde H2O. S280.This in turn led to a more complete interpretation and Burnham, C. W., 1979. The importance of volatile constituents. In:

Yoder. H. S. (ed.) The Evolution of the Igneous Rocks. Princeton, NJ:understanding of the reaction history of the Fall MountainPrinceton University Press, pp. 439–482.nappe. We now conclude that much of the metamorphism

Burton, K. W., Kohn, M. J., Cohen, A. S. & O’Nions, R. K., 1995.was accompanied by little if any fluid infiltration, andThe relative diffusion of Pb, Nd, Sr and O in garnet. Earth andfinal juxtaposition of the nappes probably occurred afterPlanetary Science Letters 133, 199–211.

substantial cooling. Questions concerning the late hy- Chamberlain, C. P., 1985. Tectonic and metamorphic history of adration of the nappe and the production of Grt5 are high-grade terrane, southwestern New Hampshire. Ph.D. Thesis,not completely resolved, but are clearly linked to the Harvard University, Cambridge, MA.

Chamberlain, C. P., 1986. Evidence for the repeated folding of iso-mechanisms and kinetics of fluid flow and mass transport,therms during regional metamorphism. Journal of Petrology 27, 63–89.as they in turn are linked to rock fabrics and the physical

Chamberlain, C. P. & Lyons, J. B., 1983. Pressure, temperature andand chemical characteristics of constituent minerals. Formetamorphic zonation studies of pelitic schists in the Merrimackexample, the anisotropic growth of Grt5 parallel to theSynclinorum, south–central New Hampshire. American Mineralogist

foliation may reflect growth in a differential stress field, 68, 530–540.but may also indicate that late retrograde mass transport Coghlan, R. A. N., 1990. Studies in diffusional transport: grain bound-across the contact had a large advective component along ary transport of oxygen in feldspars, diffusion of oxygen, strontium,the foliation. If the latter is true, longitudinal permeability and the REE’s in garnet, and thermal histories of granitic intrusions

in south–central Maine using oxygen isotopes. Ph.D. Thesis, Brownfar exceeded transverse permeability, as has been in-University, Providence, RI.dicated in several previous studies (e.g. Rumble & Spear,

Elsenheimer, D. & Valley, J. W., 1993. Sub-millimeter scale zonation1983; Ferry, 1988, 1994; Holdaway & Goodge, 1990;of d18O in quartz and feldspar, Isle of Skye, Scotland. Geochimica etGerdes & Valley, 1994; Kohn & Valley, 1994; BowmanCosmochimica Acta 57, 3669–3676.

et al., 1994; Goodge & Holdaway, 1995; Gerdes et al., Farver, J. R. & Yund, R. A., 1991. Oxygen diffusion in quartz:1995; Skelton et al., 1995). dependence on temperature and water fugacity. Chemical Geology 90,

55–70.Ferry, J. M., 1988. Contrasting mechanisms of fluid flow through

adjacent stratigraphic units during regional metamorphism, south–ACKNOWLEDGEMENTS central Maine, USA. Contributions to Mineralogy and Petrology 98, 1–12.

Ferry, J. M., 1994. Overview of the petrologic record of fluid flowWe thank M. Spicuzza for maintaining the laser ex-during regional metamorphism in northern New England. Americantraction line, J. Fournelle for his help with the electronJournal of Science 294, 905–988.microprobe, and B. Hess for preparing special thick

Fortier, S. M. & Giletti, B. J., 1989. An empirical model for predictingsections for oxygen isotope analysis. We also gratefullydiffusion coefficients in silicate minerals. Science 245, 1481–1484.acknowledge the Albert and Alice Weeks Visiting Dis-

Fortier, S. M. & Giletti, B. J., 1991. Volume self-diffusion of oxygentinguished Professorship, from the Department of Geo- in biotite, muscovite, and phlogopite micas. Geochimica et Cosmochimicalogy and Geophysics, University of Wisconsin, for Acta 55, 1319–1330.supporting F.S.S. during his term at UW. This paper Gardien, V., Thompson, A. B., Grujic, D. & Ulmer, P., 1995. Ex-

perimental melting of biotite+plagioclase+quartz±muscovite as-substantially benefited from excellent, detailed commentssemblages and implications for crustal melting. Journal of Geophysicalfrom Sorena Sorensen, John Goodge, John Ferry, andResearch 100, 15581–15591.John Bowman. John Goodge is thanked for emphasizing

Gerdes, M. L. & Valley, J. W., 1994. Fluid flow and mass transportthe importance of stress fields. This work was funded byat the Valentine wollastonite deposit, Adirondack Mountains, New

NSF Grants EAR 9316349 (M.J.K.), EAR 9220094 York State. Journal of Metamorphic Geology 12, 589–608.(F.S.S.), and EAR 9304372 ( J.W.V.), DOE grant FG02- Gerdes, M. L., Baumgartner, L. P. & Person, M., 1995. Stochastic93ER14389 ( J.W.V.), and an NSF postdoctoral fel- permeability models of fluid flow during contact metamorphism.

Geology 23, 945–948.lowship to M.J.K.

1271


Giletti, B. J. & Yund, R. A., 1984. Oxygen diffusion in quartz. Journal (eds) Guidebook for Field Trips in the Connecticut Valley Region of Massachusetts

and Adjacent States. Amherst, MA: Dept. of Geology and Geography,of Geophysical Research 29, 4039–4046.Univ. of Massachusetts, pp. 446–466.Goodge, J. W. & Holdaway, M. J., 1995. Rock-buffered fluid evolution

Spear, F. S. & Florence, F. P., 1992. Thermobarometry in granulites:of metapelites and quartzites in the Picuris Range, northern Newpitfalls and new approaches. Journal of Precambrian Research 55, 209–Mexico. Journal of Petrology 36, 1229–1250.241.Holdaway, M. J., 1971. Stability of andalusite and the aluminum

Spear, F. S. & Kohn, M. J., 1996. Trace element zoning in garnet assilicate phase diagram. American Journal of Science 271, 97–131.a monitor of crustal melting. Geology 24, 1099–1102.Holdaway, M. J. & Goodge, J. W., 1990. Rock pressure vs. fluid

Spear, F. S., Hickmott, D. D. & Selverstone, J., 1990a. Metamorphicpressure as a controlling influence on mineral stability: an exampleconsequences of thrust emplacement, Fall Mountain, New Hamp-from New Mexico. American Mineralogist 75, 1043–1058.shire. Geological Society of America Bulletin 102, 1344–1360.Kerrick, D. M., 1972. Experimental determination of

Spear, F. S., Kohn, M. J., Florence, F. & Menard, T., 1990b. A modelmuscovite+quartz stability with P H20<P total. American Journal of Sciencefor garnet and plagioclase growth in pelitic schists: implications for272, 946–958.thermobarometry and P–T path determinations. Journal of Meta-Kohn, M. J., 1993. Modeling of prograde mineral d18O changes inmorphic Geology 8, 683–696.metamorphic systems. Contributions to Mineralogy and Petrology 113,

Spear, F. S., Kohn, M. J. & Paetzold, S., 1995. Petrology of the24–39.regional sillimanite zone, west–central New Hampshire, U.S.A., withKohn, M. J. & Valley, J. W., 1994. Oxygen isotope constraints onimplications for the development of inverted isograds. Americanmetamorphic fluid flow, Townshend Dam, Vermont, USA. Ge-Mineralogist 80, 361–376.ochimica et Cosmochimica Acta 58, 5551–5566.

Thompson, A. B. & Algor, J. R., 1977. Model systems for anatexis ofKohn, M. J. & Valley, J. W., 1997. Obtaining equilibrium oxygenpelitic rocks I. Theory of melting reactions in the system KAlO2–isotope fractionations from rocks: theory and examples. ContributionsNaAlO2–Al2O3–SiO2–H2O. Contributions to Mineralogy and Petrology 63,to Mineralogy and Petrology (in press).247–269.Kohn, M. J., Valley, J. W., Elsenheimer, D. & Spicuzza, M. J., 1993.

Thompson, A. B. & Tracy, R. J., 1979. Model systems for anatexis ofOxygen isotope zoning in garnet and staurolite: evidence for closedpelitic rocks. Contributions to Mineralogy and Petrology 70, 429–438.system mineral growth during regional metamorphism. American

Thompson, J. B., Jr & Norton, S. A., 1968. Paleozoic regional meta-Mineralogist 78, 988–1001.morphism in New England and adjacent areas. In: Zen, E.-A.,Kretz, R., 1983. Symbols for rock-forming minerals. American MineralogistWhite, W. S., Hadley, J. B. & Thompson, J. B., Jr (eds) Studies of68, 277–279.Appalachian Geology, Northern and Maritime. New York: John Wiley, pp.Kruger, F. C., 1946. Structure and metamorphism of the Bellows Falls319–327.Quadrangle of New Hampshire and Vermont. Geological Society of

Thompson, J. B., Jr. & Rosenfeld, J. L. R., 1979. Reinterpretation ofAmerica Bulletin 57, 161–206.nappes in the Bellows Falls–Brattleboro area, New Hampshire–Le Breton, N. & Thompson, A. B., 1988. Fluid-absent (dehydration)Vermont. In: Skehan, J. W. & Osberg, P. H. (eds) The Caledonides inmelting of biotite in metapelites in the early stages of crustal anatexis.the USA—Geological Excursions in the Northeast Appalachians. Weston,Contributions to Mineralogy and Petrology 99, 226–237.MA: Weston Observatory, pp. 117–121.McGuire, A. V., Francis, C. A. & Dyar, M. D., 1992. Mineral standards

Thompson, J. B., Jr, Robinson, P., Clifford, T. N. & Trask, N. J.,for electron microprobe analysis of oxygen. American Mineralogist 77,1968. Nappes and gneiss domes in west–central New England. In:1087–1091.Zen, E.-A., White, W. S., Hadley, J. B. & Thompson, J. B., Jr (eds)McLellan, E. L., 1983. Contrasting textures in metamorphic andStudies of Appalachian Geology, Northern and Maritime. New York: Johnanatectic migmatites: an example from the Scottish Caledonides.Wiley, pp. 203–218.

Journal of Metamorphic Geology 1, 241–262.Thompson, P. J., 1985. Stratigraphy and structure of the MonadnockMcLellan, E. L., 1989. Sequential formation of subsolidus and anatectic

quadrangle, New Hampshire. Ph.D. Thesis, University of Mas-migmatites in response to thermal evolution, eastern Scotland. Journal

sachusetts, Amherst.of Geology 97, 165–182.

Tracy, R. J., 1978. High grade metamorphic reactions and partialOlsen, S. N., 1987. The compositions and role of the fluid in migmatites:

melting in pelitic schist, west–central Massachusetts. American Journala fluid inclusion study of the Front Range rocks. Contributions to

of Science 278, 150–178.Mineralogy and Petrology 96, 104–120. Valley, J. W., Kitchen, N., Kohn, M. J., Niendorf, C. R. & Spicuzza,

Patino-Douce, A. E. & Johnston, A. D., 1991. Phase equilibria and M. J., 1995. UWG-2, a garnet standard for oxygen isotope ratios:melt productivity in the pelitic system: implications for the origin of strategies for high precision and accuracy with laser heating.peraluminous granitoids and aluminous granulites. Contributions to

Geochimica et Cosmochimica Acta 59, 5223–5231.Mineralogy and Petrology 107, 202–218. Vielzeuf, D. & Holloway, J. R., 1988. Experimental determination of

Rosenfeld, J. L., 1969. Stress effects around quartz inclusions in the fluid-absent melting relations in the pelitic system: consequencesalmandine and the piezothermometry of coexisting aluminum sil- for crustal differentiation. Contributions to Mineralogy and Petrology 98,icates. American Journal of Science 267, 317–351. 257–276.

Rumble, D., III & Spear, F. S., 1983. Oxygen-isotope equilibration and Young, E. D., 1993. On the 18O/16O record of reaction progress inpermeability enhancement during regional metamorphism. Journal of open and closed metamorphic systems. Earth and Planetary Sciencethe Geological Society, London 140, 619–628. Letters 117, 147–167.

Skelton, A. D. L., Graham, C. M. & Bickle, M. J., 1995. Lithologicaland structural controls on regional 3-D fluid flow patterns duringgreenschist facies metamorphism of the Dalradian of the SW ScottishHighlands. Journal of Petrology 36, 563–586. APPENDIX A: THE ORIGIN OF Grt5Spear, F. S., 1992. Inverted metamorphism, P–T paths and cooling

The most perplexing feature of Grt5 is their chemicalhistory of west–central New Hampshire: implications for the tectonicevolution of central New England. In: Robinson, P. & Brady, J. B. disparity compared with Grt4, especially concerning Ca

1272


systematics. This disparity can only result from two (3) Grt5 is well developed only in rocks that containboth chlorite and sillimanite.general processes: (1) changes of P–T that affect element

A disadvantage of the model is that there is no ex-partitioning, or (2) transient changes to the mass balancetremely calcic plagioclase, but that component wouldof the rock (i.e. open-system and/or limited mass trans-also be the first destroyed during garnet growth.port effects). Several observations suggest that Grt5 is not

simply the result of changes in P or T in a closed system.First, Grt5 is texturally associated with chlorite, whichrequires infiltration of hydrous fluids. Second, limits on

APPENDIX B: ANALYTICALpossible P–T paths can be calculated based on the chem-ical zoning of Grt5 and the range of zoning observed in TECHNIQUESmatrix plagioclase, and demonstrate that in a closed Electron microprobe analyses (Table 1) were collectedsystem garnet cannot grow with the observed chemical with a fully automated Cameca SX-50 in the Departmentzoning. For the chemical system MnO– of Geology and Geophysics, University of Wisconsin.Na2O–CaO–K2O–FeO–MgO–Al2O3–SiO2–H2O and Standards used included: San Carlos olivine (Si, Mg),the most likely assemblage(s), Grt+Bt+Sil+Qtz+ Rockport fayalite (Fe), Great Sitkin anorthite (Ca), Ame-Pl+Ms±H2O, these calculated paths all imply a strong lia albite (Na), synthetic F-phlogopite (K, F), and naturaldecrease in pressure at nearly constant temperature, and rhodonite (Mn), sillimanite [Al, No. 131013 of McGuire

et al. (1992)], rutile (Ti), and almandine [O, No. 112140 ofuniformly consume rather than produce garnet.McGuire et al. (1992)]. Operating conditions for completeInfiltration-driven formation of Grt5 does explain theanalyses were a 15 kV accelerating voltage and a 20 nAgarnet chemical variations, as illustrated with a relativelyflag current. The beam was slightly defocused to 5 lmsimple, albeit non-unique interpretation of Grt5 growth.for analyses of white mica and feldspars, and all raw dataInitial infiltration of H2O into a chlorite-absent as-were reduced using a Uqz correction scheme. Maximumsemblage would allow the reaction Alb+Sil+H2O=counting times were 20 s. X-ray maps were collected forPg+Qtz to proceed. If only a small fraction of theMg, Fe, Mn, Ca, and O (all WDS), with a flag currentplagioclase participated in the reaction, then it (andof 150–200 nA, a beam size of 2 lm, and count timescoexisting garnet) would rapidly become fairly calcic (e.g.of 30–35 ms. Operating conditions for trace element X-An70), and the small amount of paragonite producedray maps were a beam size of 5 lm, a flag current ofwould dissolve as a component in muscovite. Continued2000 nA, count times of 100 ms, and step sizes of 7 orinfiltration of H2O would then stabilize chlorite, leading16 lm.to the metastable reaction Sil+Bt+H2O=

Oxygen isotope analyses (Table 2) were collected byGrt5+Chl+Ms+Qtz. In this assemblage, garnet is pre-using the laser probe extraction line at the Departmentdicted to have decreasing Mn and increasing Fe/of Geology and Geophysics, University of Wisconsin,(Fe+Mg), as observed. If the first-formed Grt5 equi-using a CO2 laser, BrF5 reagent, and a Finnigan-MATlibrated with anorthite-rich plagioclase rims, then it would251 mass spectrometer (Elsenheimer & Valley, 1993;be fairly calcic, whereas fractional crystallization wouldKohn et al., 1993; Valley et al., 1995), and standardizedlead to the decrease in grossular towards the rim (e.g.against garnet standards UW GMGrt No. 1 and UWG-Spear et al., 1990b). 2 (Valley et al., 1995; Kohn & Valley, 1997). Mineral

Because garnet and plagioclase compositions cannot separates were prepared by crushing 5–10 g of eachbe directly correlated for Grt5 growth, it is impossible to sample, sizing between 150 and 300 lm, and hand-evaluate this scenario quantitatively and constrain any picking. No attempt was made to separate leucosomesP–T changes, but insofar as minerals equilibrate during from matrix material. Garnet and sillimanite zoninginfiltration, retrieved rim P–T conditions should be ac- studies were conducted by using the thin sawblade ap-curate. The model accommodates a simple P–T history, proach (Elsenheimer & Valley, 1993; Kohn et al., 1993),in that nearly isobaric cooling with hydrous infiltration which involves dissection of specially prepared, 600 lmand differential reaction can explain all the data, and thick wafers of each garnet and sillimanite. Analyticalfurther allows the following observations to be explained: reproducibility based on multiple standard analyses and

(1) Preferential growth of Grt5 rims within the foliation duplications of unknowns is routinely ±0·08‰.rather than across it could reflect enhanced mass transport The locations of four samples of Bethlehem Gneiss areparallel to the foliation, as might be expected for reactions imprecisely known, but can be constrained from fabrics.driven by infiltrating fluids. A very mild solid-state shear fabric is present in the

(2) Different areas of the outcrop have different degrees Bethlehem Gneiss within 3–5 m of the contact. Theof rehydration (Chl and St abundance) and Grt5 pro- equigranular texture in the four samples suggests theyduction, as expected if retrograde hydrous fluids were were collected at least 3 m from the contact, and we

have used a plotting position of 5 m.heterogeneously distributed.

1273


Table 1: Electron microprobe analyses of minerals from the Fall Mountain nappe, New Hampshire

Garnet

K92-12b2 K92-12b2 K92-12b2 K92-12b2 K92-12d K92-12d K92-12d K92-12d

Rim Max Ca Max Mn Core Core Max Mn Max Ca Rim

SiO2 36·75 37·07 36·87 36·78 36·73 36·44 36·13 36·17

Al2O3 21·04 21·10 21·10 21·12 21·12 21·00 20·80 20·84

TiO2 0·00 0·00 0·00 0·00 0·08 0·09 0·07 0·08

MgO 1·87 1·76 2·49 3·22 2·70 2·19 1·63 1·45

FeO∗ 35·32 30·33 32·29 33·85 37·52 36·59 37·56 37·71

MnO 3·54 4·74 7·78 4·31 2·05 3·27 2·07 2·31

CaO 1·91 5·86 0·93 1·22 1·08 1·07 2·30 1·85

Total 100·43 100·86 101·46 100·49 101·28 100·64 100·43 101·19

Si 2·978 2·970 2·962 2·963 2·954 2·956 2·946 2·953

Al 2·010 1·993 1·999 2·006 2·002 2·009 1·999 2·006

Ti 0·000 0·000 0·000 0·000 0·005 0·005 0·004 0·005

Mg 0·226 0·211 0·299 0·386 0·324 0·264 0·197 0·177

Fe2+ 2·394 2·033 2·169 2·281 2·523 2·482 2·561 2·575

Mn 0·243 0·322 0·530 0·294 0·140 0·225 0·143 0·160

Ca 0·166 0·503 0·080 0·105 0·093 0·093 0·201 0·162

FM 0·914 0·904 0·879 0·855 0·886 0·904 0·928 0·936

Prp 0·075 0·079 0·097 0·126 0·105 0·086 0·064 0·058

Alm 0·791 0·745 0·705 0·744 0·819 0·810 0·826 0·838

Sps 0·080 0·071 0·172 0·096 0·045 0·073 0·046 0·052

Grs 0·055 0·105 0·026 0·034 0·030 0·030 0·065 0·053

K92-12h K92-12h K92-12h K92-12I K92-12I K92–12I K92–12I

Core Max Ca Rim Core Max Mn Max Ca Rim

SiO2 36·70 36·42 36·57 36·74 36·64 36·74 36·73

Al2O3 21·25 21·14 21·27 20·90 20·64 20·71 20·80

TiO2 0·04 0·13 0·02 0·00 0·01 0·03 0·01

MgO 2·49 1·88 1·63 3·18 2·55 1·97 2·47

FeO∗ 30·99 33·94 35·39 32·63 32·47 32·53 32·55

MnO 8·87 4·58 4·54 5·40 6·75 4·75 6·59

CaO 0·85 2·86 1·45 0·99 0·80 3·42 0·76

Total 100·95 100·56 100·88 99·84 99·85 100·15 99·91

Si 2·954 2·942 2·960 2·977 2·985 2·979 2·987

Al 2·016 2·013 2·031 1·997 1·982 1·980 1·994

Ti 0·002 0·008 0·002 0·000 0·000 0·002 0·001

Mg 0·299 0·226 0·197 0·383 0·309 0·238 0·299

Fe2+ 2·086 2·293 2·396 2·211 2·212 2·206 2·214

Mn 0·604 0·314 0·312 0·371 0·466 0·326 0·454

Ca 0·073 0·247 0·126 0·086 0·070 0·298 0·066

FM 0·875 0·910 0·924 0·852 0·877 0·903 0·881

Prp 0·098 0·073 0·065 0·126 0·101 0·078 0·099

Alm 0·681 0·744 0·791 0·725 0·724 0·719 0·730

Sps 0·197 0·102 0·103 0·121 0·152 0·106 0·150

Grs 0·024 0·080 0·042 0·028 0·023 0·097 0·022

1274


Table 1: continued

Mica

K92-12b2 K92-12b2 K92-12d K92-12d K92-12h K92-12h K92-12I K92-12I K92-12I

Matrix Bt Matrix Ms Matrix Bt Matrix Ms Matrix Bt Matrix Ms Lo-Ti Bt Hi-Ti Bt Matrix Ms

SiO2 35·39 46·53 34·47 45·53 35·07 45·69 35·35 34·60 45·46

Al2O3 19·76 36·04 19·74 37·07 19·85 36·17 19·34 18·96 36·72

TiO2 1·50 0·36 1·64 0·45 1·38 0·71 1·46 2·66 0·28

MgO 9·58 0·58 7·54 0·37 9·50 0·64 10·03 8·94 0·42

FeO∗ 20·57 0·94 24·05 1·08 21·12 1·18 19·92 20·39 1·21

MnO 0·05 0·00 0·05 0·03 0·12 0·02 0·10 0·08 0·03

CaO 0·01 0·00 0·02 0·02 0·00 0·03 0·00 0·00 0·00

Na2O 0·37 1·35 0·30 1·25 0·39 1·11 0·36 0·39 1·62

K2O 9·26 10·12 8·99 9·95 8·72 10·03 9·17 8·81 9·67

Total 96·48 95·92 96·80 95·75 96·16 95·59 95·74 96·04 94·23

Si 2·671 3·069 2·637 3·011 2·657 3·029 2·682 2·658 3·018

Aliv 1·329 0·931 1·363 0·989 1·343 0·971 1·318 1·342 0·982

Alvi 0·428 1·872 0·418 1·900 0·430 1·857 0·411 0·375 1·892

Ti 0·085 0·018 0·094 0·023 0·079 0·035 0·083 0·154 0·014

Mg 1·077 0·057 0·860 0·036 1·073 0·063 1·134 1·024 0·042

Fe2+ 1·298 0·052 1·539 0·060 1·338 0·066 1·264 1·310 0·067

Mn 0·003 0·000 0·003 0·002 0·008 0·001 0·007 0·005 0·002

Sum Oct 2·892 1·999 2·915 2·021 2·928 2·022 2·900 2·868 2·017

Ca 0·001 0·000 0·002 0·001 0·000 0·002 0·000 0·000 0·000

Na 0·053 0·173 0·045 0·160 0·057 0·143 0·052 0·058 0·209

K 0·891 0·851 0·877 0·840 0·843 0·848 0·888 0·864 0·819

Sum A 0·946 1·024 0·923 1·001 0·900 0·993 0·940 0·922 1·028

FM 0·547 0·476 0·641 0·621 0·555 0·511 0·527 0·561 0·618

Chlorite Feldspar

12b2 12d 12I 12b2 Rim 12b2 Core 12d Rim 12d Core

SiO2 24·65 23·44 24·09 An21 An38 An22 An40

Al2O3 23·01 23·12 23·31

TiO2 0·05 0·18 0·08 12h Rim 12h Core 12I Rim 12I Core

MgO 14·08 11·90 14·06An14 An17 An20 An35

FeO∗ 26·41 30·60 26·40

MnO 0·11 0·07 0·06

Total 88·30 89·31 87·98

Si 2·590 2·496 2·543

Al 2·850 2·904 2·901

Ti 0·004 0·015 0·006

Mg 2·205 1·889 2·212

Fe2+ 2·321 2·726 2·331

Mn 0·009 0·007 0·006

FM 0·513 0·591 0·513

FM=Fe2+/(Mg+Fe2+).

1275


Table 2: Oxygen isotope compositions∗ of minerals from the Fall Mountain nappe, New Hampshire Bulk

separates

Rangeley Formation

Sample† Grt (1–3)‡ Grt (4–5)‡ Qtz Sil‡ Ms Bt Other‡

K95-18F 13·20,13·06 13·17,13·16 17·39 14·05,14·16 12·01,11·80

K95-18E 13·20,13·38, 13·15 17·15,17·44 14·16,14·19 11·64,11·74

K95-18D 12·98,12·94 12·97,12·67 17·18,17·10 13·97,14·02 13·81,13·89 11·97,11·85

K95-18C 13·23,13·31 13·24 17·48,17·22 14·35,14·33 14·16,14·22 12·02,12·18

K95-18B 12·19,12·30 11·99 16·66 13·22,12·99, 9·10,9·06

13·21

K95-18A 12·75,13·04 12·89,12·70 16·62,16·67 13·99,14·05 13·60,13·89

K92-12H 12·99 12·51,12·38 16·18,16·32 13·95,14·57, 10·47,10·42 Grt:12·83,13·04,12·88,

14·58,14·10, 12·81,12·92,12·92

14·19,13·85 Fibr: 14·29

K92-12I 12·91,12·91 16·34,16·39 14·01,14·23, 11·52,11·31 Grt: 13·04,13·19,13·05

13·81,14·00 Fibr: 13·76

K92-12G 13·01 12·73 16·03,16·12 13·94,13·95 10·35,10·72 St: 13·05,13·00

K92-12J 12·28,12·32 11·84,11·73,11·38 15·62,15·64 13·06 9·89 St: 11·63

K92-12E 12·09,12·07 11·80 15·70 13·20,13·28 9·87,10·11 St: 11·91,11·75

K92-12D(1) 13·98,14·24 8·54,8·87

K92-12D(2) 14·74,14·49 12·34 9·22,9·28 Grt: 10·89

K92-12D(3) 11·45,11·76 11·11 14·70,14·70 8·49,8·67 Grt: 11·27,11·45

K92-12D(4) 11·82,11·86 11·00,11·17 14·53,14·56 8·88,8·41 Grt: 11·58,11·58

K92-12D(5) 14·62,14·57 8·90,8·11

K92-12B 12·71 15·45,15·70 12·94,13·03, 9·88,10·11 St: 12·79,12·51

12·97,13·10 Fibr: 12·95

Bethlehem Gneiss

Sample† Grt Fsp Qtz Sil Ms Bt

K92-12M 11·77,11·74 14·37,14·51 11·30,11·29 9·03,9·17

K92-12N 9·95 12·08,11·97 14·37,14·49 11·54,11·34 9·38,9·07

K92-12P 9·71,9·85 11·65,11·91 14·03,13·74 10·87,10·86 8·22,8·28

BF-10A 9·98,10·05 10·93,10·94 14·02,14·14 10·85,10·87 7·33,7·36

BF-10B 12·57,13·22 14·38,14·61 7·23,7·30 8·49,8·50

BF-13D1 9·80,9·73 10·67,10·65 13·52,13·85 8·42,8·32,8·45

BF-13D2 9·67,9·67 10·59,10·89 13·70,13·51,13·45 8·28,8·24

1276


Table 2: continued

Sample & Min’l Dist§ d18O Dist d18O Dist d18O Dist d18O Dist d18O Dist d18O

K92-12D Grt 0·1 11·15 0·1 11·31 0·2 11·59 0·2 11·68 0·25 11·46 0·25 11·36

(5·0 mm diam.) 0·25 11·61 0·3 11·88 0·3 11·62 0·3 11·33 0·4 11·76 0·5 12·18

0·7 12·31 0·8 12·04 0·8 12·28 1·2 12·39 1·2 11·92 1·5 12·52

1·5 12·47 1·5 12·36 1·5 12·30 1·6 12·14 1·8 12·41 2·0 12·26

2·3 12·31 2·4 12·27 2·4 11·90

K92-12A Grt 0·4 10·46 1·1 10·63 1·8 10·62 2·4 10·82 2·7 10·57 3·7 10·63

(10·2 mm diam.) 4·9 10·58 4·2 10·76 3·4 10·52 2·7 10·86 1·8 10·52 0·9 10·60

0·2 10·65

K92-12C Grt 0·4 11·71 0·5 11·65 0·6 11·72 0·7 12·18 1·0 11·93 1·2 11·67

(6·0 mm diam.) 1·3 11·93 2·0 11·74

K92-12J Sil(x) 0·4 13·60 1·0 13·41 0·4 13·50

(2·0×8·5 mm)

K92-12J Grt 0·8 12·96 0·3 12·73

(1·5 mm diam.)

K92-12B Sil(l) 0·3 13·56 2·2 13·52 2·7 13·75 2·1 13·65 1·6 13·53 1·0 13·49

(5·6×>7 mm)

K92-12H Sil (l) 2·3 14·29 2·6 14·19 3·2 14·25 4·6 13·83 4·6 13·97 6·3 14·09

(2·8×13·3 mm) 6·5 13·96 5·5 14·20 4·1 14·28 4·1 14·18 2·9 14·12 2·1 14·17

1·8 14·12

K92-12H Sil (l) 0·6 14·39 1·3 14·25 1·9 14·14 2·5 14·36 3·5 14·32 3·4 14·11

(1·2×8·0 mm) 2·7 14·69 1·8 14·55

K92-12I Sil (l) 0·5 13·83 1·2 13·76 1·7 13·87 0·9 13·94 0·3 13·92

(1·2×3·9 mm)

∗All analyses are in ‰ relative to V-SMOW, and were standardized against the garnet standard UWG-2 and UW GMGrt No.1, assuming values of 5·8‰ and 6·2‰, respectively (Valley et al., 1995; Kohn & Valley, 1997).†Samples K92-12 are from the southern exposure, whereas samples K95-18 are from the northern. Distances from thecontact (in meters) perpendicular to the foliation are: K92-12P, –1·75; K92-12N, –0·2; K92-12M, –0·2; K92-12A, 0·5; K92-12B,0·5; K92-12C, 0·5; K92-12D, 0·5; K92-12E, 0·5; K92-12J, 6; K92-12G, 10; K92-12I, 12; K92-12H, 13·5; K95-18A, 15; K95-18B, 17;K95-18C, 25; K95-18D, 35; K95-18E, 45; K95-18F, 55. Negative distances are below the contact (i.e. within the BethlehemGneiss), and positive distances are above the contact in the Rangeley Formation. Samples BF-10A, BF-10B, BF-13D1, andBF-13D2 were not specifically located, but were probably collected at least 5 m below the contact. Samples K92-12D(1)–(5)are sublayers of ~1 cm thickness from sample K92-12D. Subdivision 1 contains the coarse garnets that were used todetermine intragranular isotope zoning trends, and is the furthest sublayer from the contact with the Bethlehem Gneiss.Relative modes of Qtz/Fsp/Bt/Ms for the Bethlehem Gneiss samples (to nearest 5%) were: K92-12M, 20/50/20/10; K92-12N,20/55/20/5; K92-12P, 40/40/5/15; BF-10A, 30/40/25/5; BF-10B, 20/60/15/5; BF-13D1, 20/50/30/0; BF-13D2, 25/55/20/0.‡‘Grt(1–3)’, first population (pink garnet) separates, which are interpreted to be principally second-generation garnet withsome first- and third-generation garnet; ‘Grt(4–5)’, second population (orange garnet) separates, which are interpreted tobe garnet generations 4–5; ‘Grt’, no differentiation made among garnet populations; ‘Sil’, prismatic sillimanite; ‘Fibr.’,fibrolitic sillimanite. Other mineral abbreviations are from Kretz (1983).§‘Dist’ indicates distance to the nearest grain boundary in mm. Sillimanite dimensions reflect width and length; ‘(x)’ and‘(l)’ indicate whether the traverse was perpendicular or parallel to the long dimension.

1277

Date post:	20-May-2018
Category:	Documents
Upload:	vanmien
View:	216 times
Download:	2 times

Dehydration-Melting and Fluid Recycling during...

Documents